damage and huge current draw from Teensy 3.2

Status
Not open for further replies.

steve_o

New member
Hi all, longtime Teensy fan here and first time poster. After 3+ years of Teensy-related triumphs it appears I've just fried a few Teensy 3.2s for reasons which are beyond me, and I was hoping some folks here might offer their thoughts.

I have a circuit in which a Teensy 3.2 is controlling a motor driver and a few simple digital and analog I/Os (just a few buttons and knobs). I'm running the Teensy and the motor driver off the same 5V 1A switching power regulator. After running the circuit with the motor spinning for anywhere from a few minutes to a few hours, the Teensy will suddenly get hot, and if left on for long enough after that (a few minutes) will simply die from the heat. But if I feel the Teensy getting hot at the moment it first starts heating up and power down til it cools off, it'll still run (and still be programmable) until it gets hot again after a few minutes and I have to power down again. Once it happens once, the heat continues any time I power it up, even if I put an empty program on it and/or completely remove it from the circuit besides Vin and ground. Also worth noting is that nothing else in the circuit has gotten warm or been damaged in any way.

So here's the one clue I've been able to find: If I measure the current into the Vin pin when I first start testing the circuit and program with a brand new Teensy, the Teensy is only pulling ~20mA total. No matter what I do with the motor driver and I/Os, it stays right around there, where it should. Once it inevitably starts getting hot its suddenly pulling ~4-500mA, even if I completely remove the Teensy from the circuit and only leave power and ground connected. This same thing has happened with a handful of Teensy 3.2s over the past few weeks, the only difference being one damaged one ended up pulling closer to 250mA rather than 4-500mA.

I've been staring at my scope to see if repeatedly starting and stopping the motor is causing a potentially lethal spike in voltage, or a brownout, and it seems to stay good and steady at 5v as far as I can tell even when there's a big spike in current on the motor when being switched on and off.

So I'm having a hard time diagnosing, and I know it would be difficult to impossible to diagnose from afar. Can anyone tell me, does the massive current draw from the Teensy I'm seeing at least point to any specific type of damage? Any other thoughts? Any responses are much appreciated, thank you.
 
Voltage spikes can damage a chip like that so fast it's hard to see even on a decent oscilloscope. It's already blown so why not measure the resistance to both ground and 3.3v on all the in pins you use to maybe see if one of them is blown and to give an idea of it it came in through an io on. My guess is it didn't. If it came in on 5v did you have a reversed biased diod across the motor? A cap or series resistor and cap across it? These can help but separate power sources are safer and opto couplers to isolate the drive circuits completely add more safety. There are tvs like this http://www.mouser.com/Search/m_Prod...GAEpiMZZMuNo3spt1BaV1%2bwaeMMeYLm0VVex/4ylX0= that can help. It's not recommended anywhere I can find but I like putting a small resistance in series with the 5 v and one of those installed like a zener diode on the teensy side to ground. It can be a ohm or less. I want it to burn before anything else. A 1/4w would do as max teensy current is under 1/2 amp.
 
I forgot to mention in another thread some one suggested putting 330 or 470 ohm resistors in series with your output pins. That's an excellent suggestion too. I've seen "industrial" Arduinos clones that do that. They only cause issues if what your driving isn't high impedance or needs level changes driven very fast. To avoid killing another teensy (R.IP.) consider overkill and include suggestions from lots of people and include several of them.
 
Last edited:
Consider protection diodes too

Our hardware needs to be protected from all sorts of environmental insults (spikes, etc.) so we use protective diodes. The part we use is a three-terminal part so it fits in a very tightly packed board. With the diodes in place, the spike goes toward the power supply (overshoot) or ground (undershoot), mostly preventing backfeed into the signal pin. With the diodes in place, we are more comfortable with 10-100 ohm resistors in cases where the diodes are not sufficient.

esd-protection-diodes-circuit.gif
 
Sorry if its bad form to revive an old thread but I just wanted to give a big thank you to both of you for responding!

I spent some time staring at a scope while jerking the motor around as hard as possible and was able to see extremely brief (nanos) but huge spikes (and occasionally drops) in voltage on the Teensy Vin pin that corresponded 100% with Teensy getting hot.

I've since been powering Teensy through a 1n400x diode to prevent negative voltage, as well as a TVS diode from power to ground with a breakdown of 6.1v and a clamping voltage of 6.5v. If I understand correctly any voltage over 6v can damage a Teensy due to heat even though the ICs are technically rated for 10v (right?), but this was the closest I could find with a working voltage around 5v. After hours of testing, so far its eliminated the spikes, under the conditions which previously causes the spikes.

So hopefully this is it! I know using a separate power supply is technically the best solution, but the rest of the circuit and the enclosure is already designed so hopefully the added diode protection will prove bulletproof.

Thanks again for your help!
 
Status
Not open for further replies.
Back
Top